INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ĸļ

-0.77

rooms

-0.72

 mattered

-0.66

 royalty

-0.65

room

-0.64

 elapsed

-0.63

 Seym

-0.63

embed

-0.63

ÙĴ

-0.61

âĶĢâĶĢâĶĢâĶĢ

-0.61

POSITIVE LOGITS

essor

0.78

 Robotics

0.66

umenthal

0.66

essors

0.65

imir

0.64

inson

0.63

heric

0.63

ospace

0.63

othe

0.62

igl

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.